Ovarian cancer remains a common and lethal disease. Most women diagnosed with ovarian cancer are destined to die from advanced, refractory disease, even though most also respond well to initial therapy. At least 80% of human cancers depend on aberrant signaling from receptor tyrosine kinases (RTK) to the mitogen-activated protein kinase (MAPK) ERK. These signals drive proliferation, survival, and expression of the malignant phenotype in cancer cells and also critically maintain tumor micro- environments. Moreover, aberrant RTK-ERK signaling via Her2/neu (ERBB2) is an adverse prognostic factor in ovarian cancer. Sprouty 2 (SPRY2) protein is a tumor suppressor and endogenous inhibitor of RTK-ERK signaling. Many of its actions depend on a dynamic equilibrium with Cbl, an E3-ligase that targets SPRY2 as well as EGF-family receptors for proteosomal degradation. SPRY2 is inactivated in 25-40% of breast, prostate, hepatocellular, lung, and melanocytic cancers. In preliminary studies, we did not detect SPRY2, by immunohistochemistry, in a similar fraction of ovarian cancer patients. Patients without detectible SPRY2 had uniformly good outcomes (all with disease-free intervals >60 mos). The central hypothesis of this application is that Spry2 predicts post-chemotherapy outcomes in advanced ovarian cancer. Two Specific Aims are proposed as initial tests of this hypothesis: (1) to determine whether the presence or level of SPRY2 predicts survival in women with advanced ovarian cancer;and, (2) to determine whether SPRY2 provides additional predictive information in context with Her2 status. We will address these Aims in 299 archival tumor specimens, originally obtained from women with advanced-stage (III-IV) ovarian carcinoma, treated on five national phase III trials (GOG 114, 132, 152, 158, and 162) In aim 1, SPRY2 and Cbl protein levels, assayed by quantitative automated immunohistochemistry, are entered into Cox models as potential predictors of survival and disease-free interval. Multivariate Cox models will include known clinical prognostic factors (grade, stage, extent of debulking, drug treatment, and performance status). In Aim 2, the ability of SPRY2 to predict clinical outcomes in context with Her2 status (amplified versus not, by FISH) will be determined using similar statistical methods. These initial studies of SPRY2, a novel candidate biomarker, will hopefully provide a basis to better predict outcomes following chemotherapy i ovarian cancer. PUBLIC HEALTH RELEVANCE: has become apparent that many cancers depend on signaling from specific proteins, called receptor tyrosine kinases (RTK) to a protein called ERK, including ovarian cancer. Ovarian cancer is a major public health problem. It is the fifth most common cause of cancer-related death in women, responsible for more deaths than all other gynecologic malignancies combined. Therapies help women, but many still relapse and succumb to their disease, We presently lack the means to predict which patient's tumors will respond to therapy and which will not. We have found that SPRY2, a key regulator of RTK to ERK signaling, is absent in about 1/3 of a small set of ovarian cancers we have investigated. We propose to test whether levels of SPRY2 predict survival of women with advanced ovarian cancer who have been treated with standard chemotherapy regimens. Information gained from these studies will help us better understand the biology of RTK function in tumors, and may be used by subsequent women and physicians to make therapeutic choices. Patients could also choose to avoid the adverse quality of life impact of ineffective therapies. By refining prediction of outcomes post-chemotherapy, the proposed studies could therefore potentially translate into significant improvements in public health.